Clustering of the inland waterway navigational environment and its effects on ship energy consumption

Abstract

This article examines the distribution characteristics of the navigational environment in the Yangtze River trunk line using several information collection sensors installed on ships that navigate in this line. Through experiments on these ships, data of energy consumption and the navigational environment are collected. Water flow and waterway depth are proved to be the main influencing factors on the ship energy consumption via Spearman’s correlation analysis. Next, data of water velocity and waterway depth that cover the entire trunk line are graded using the k-means clustering algorithm. To build an evaluation matrix of navigational environment, the frequency distribution of each grade in different Yangtze River legs is counted statistically, and on this basis, similar legs are clustered using the hierarchical clustering algorithm. In this way, the waterway partition in the Yangtze River trunk line is completed. Furthermore, the distribution of the energy consumption of ships in different legs is also calculated. The study results indicate that not only the navigational environment of the Yangtze River trunk line but also the energy consumption level of ships have distinctive regional differences. Finally, the laws of the Yangtze River navigational environment are analyzed, and the corresponding energy-saving navigation strategies are proposed, which are useful for crews to operate their ships in energy-efficient and safe conditions.